- Boost-users - lists.boost.org

[random] comments on speed?
by Sanyi 30 Jan '06

30 Jan '06

Sorry if double posted... Can anyone who have tried Boost's `random' library have any thoughts on its speed? I looked into its binomial variate function and apparently it actualy flips a (biased) coin countless times to generate it. I am not an expert in this, but this seems awfully slow for high n's. (Not to mention that I could have done it myself.) AFAIK there are fast and good algorithms/approximations for generating some non-uniform variates. Are any of these at least planned to be included? <OT> I've found some papers on fast binomial generators, but that takes some serious time and effort to translate into working code. As I don't want to reinvent the wheel, if you know of any publicly available libraries for it, let me know. </OT> Thanks, Sanyi

3 3

boost::iostream - no way to set failbit or badbit in stream from device?
by Oliver.Kowalke＠infineon.com 30 Jan '06

30 Jan '06

Hallo, how can I set failbit or badbit in boost::iostream::stream if I detect some error in boost::iostream::device::read/write? Is it possible without throwing exceptions? Example: in the write method of a device which models a socket connection an EPIPE error (peer has closed socket) was detected. How can I set the boost::iostream::stream (which uses the device) to be not useable? Regards, Oliver

1 0

Re: [Boost-users] boost::iostream - returning EOF in device::writeand set failbit
by Oliver.Kowalke＠infineon.com 30 Jan '06

30 Jan '06

Hello Petro, >> would the eofbit be set if I return EOF in device::write? >> How can I set failbit from device? >Does it make sense for write to return EOF? >EOF means "no more data". >But writing is inserting data, not requesting it. If you write to a socket and the peer has already closed the socket you get an EPIPE error. How would you propagate this to your application? Oliver

1 0

[boost][serialization] How to implement serialization classes in two DLLs
by jf332-3 29 Jan '06

29 Jan '06

Hi, After few hour on the subject, Im would like to have some explanations & recommendations one the boost serialization library. To explain, my problem, you will find hereafter a VC++ 7.1 project that demonstrates it. Problem Description We have involved in a large project with for example two DLLs libraries: LibA and LibB. These libraries define & implement one class each other, lets say libA::A and libB::B. These two classes can be serialized through boost serialization engine. The interfaces for the two classes are the same: ---------------------- struct LIBA_API A { std::vector<double> my_vector ; private: friend class boost::serialization::access; template<class Archive> void save(Archive & ar, const unsigned int version) const ; template<class Archive> void load(Archive & ar, const unsigned int version) ; BOOST_SERIALIZATION_SPLIT_MEMBER() }; LIBA_API void save(const libA::A* pObj, std::ostream& os) ; LIBA_API void load(libA::A* &pObj, std::istream& is) ; }; // Namespace libA ---------------------- The two free functions save() and load() are defined to be able to serialize the class libA::A (respectively the class libB::B). When, we try to use these interfaces in an EXE application: ---------------------- #include <LibA/A.h> #include <LibB/B.h> #include <boost/filesystem/path.hpp> #include <boost/filesystem/fstream.hpp> #pragma comment(lib, "liba_d.lib") #pragma comment(lib, "libb_d.lib") int main(int argc, char* argv[]) { libA::A a; libA::save(&a, boost::filesystem::ofstream("a.xml")); libB::B b; libB::save(&b, boost::filesystem::ofstream("b.xml")); // -> Assert in debug mode return 0; } ---------------------- The figure hereafter describes the relationships between LibA, LibB and EXE. ------------ ------------ | LibA | | LibB | ------------ ------------ \ / \ / Exe Application We have to face to an assert in debug mode generate by the file boost::serialization::extended_type_info.cpp, line#71. This assert happens during the instance serialization of class libB:B. Recently, there are some posts on this subject, but I would like to known how I can deal with this assert? Is it possible to avoid it by a code organisation? To have a concrete answer, I have joined my complete VC++7.1 project. Thanks in advance for your help. Development Platform WindowsXP VC++ 7.1 Boost 1.33.1 Marc Viala mailto:mviala@acticm.com

1 0

Re: [Boost-users] boost::serialisation - portable_binary_oarchivenotpart of library
by Oliver.Kowalke＠infineon.com 28 Jan '06

28 Jan '06

Hello Robert, I think a 'finished' protable_binary_oarchive/ protable_binary_iarchive would be very useful for network applications (passing message structures -> header, body, trailer). I'm not familiar with boost::serialization (my first look into it) - I see some specializations of load for integral types an a templated member function for all others. Serialisation for std::strings and chars are supported in a portable manner? Only floating point types have to be implemented for portability? Oliver >I made it as an example on how to extend the library by deriving from an >existing >archive. To "finish" it would require implementing serialization of >floating point >types in a portable manner and this turns out to be a non-trivial task. >Implementing >this would be too much work and would sort of compromise illustrative value >of the example.

3 3

Re: [Boost-users] boost::serialisation -Non-default Constructors
by Merrill Cornish 28 Jan '06

28 Jan '06

Yes. Each class you want to serialize needs a default constructor in addition to whatever other constructors your application needs. Merrill

2 1

::spirit, token_node_d
by elviin 28 Jan '06

28 Jan '06

Before I'll investigate more I'd like to know if anybody knows why it is not allowed a rule object inside this directive: token_node_d.

2 2

[statechart] Ready for addition to main CVS
by Andreas Huber 28 Jan '06

28 Jan '06

Dear Boosters I believe I have addressed all the important issues that were reported during formal review of the Boost.Statechart library (then proposed as boost::fsm). Although there's still a lot to do, I think that the current version is stable enough to be added to the main CVS. All reviewers (and anyone else interested) are invited to verify that the issues that have been put forth have been taken care of appropriately: http://boost-sandbox.sf.net/libs/statechart Given Thomas Witt's cutoff date for additions of new libraries (February 8th) and the roughly one week needed to move the statechart directories from the sandbox CVS to the main CVS (can only be done by source forge staff), I must ask people to come forward with their questions/concerns before February 1st. Should anyone need more time, please let me know immediately by following up to this post. Thanks & Regards, -- Andreas Huber When replying by private email, please remove the words spam and trap from the address shown in the header.

1 0

[graph] non-contiguous edges?
by Roman Schindlauer 28 Jan '06

28 Jan '06

Hello, Warning ahead - this is a newbie question. I need to find weak and strong components of a graph. So far, my code is as simple as it could be (edges being a std::vector of std::pair<unsigned,unsigned>): typedef adjacency_list <vecS, vecS, undirectedS> Graph; Graph G; for (Edges::const_iterator ei = edges.begin(); ei != edges.end();++ei) add_edge((*ei).first, (*ei).second, G); std::vector<int> component(num_vertices(G)); int num = connected_components(G, &component[0]); However, my vertices are not contiguously numbered. Hence, the following edges (8, 9) (9, 10) (12, 13) result in these connected components: 0: 0 1: 1 2: 2 3: 3 4: 4 5: 5 6: 6 7: 7 8: 8, 9, 10 9: 11 10: 12, 13 which is not what I want (just two wcc: 8, 9, 10 and 12, 13). I was reading into the property_map stuff, but as far as i understand, that would not help here. Do I have to remap the indices beforehand or is there a way to convince Boost that there are no other nodes than those in the edges? Best regards, Roman

1 0

Warnings and boost
by gast128 27 Jan '06

27 Jan '06

Dear all, Our company makes heavy use of boost libraries. But unfornately some libraries trigger warnings. This is hard to circumvent since we both have level 4 and warnings as error option enabled in Visual Studio 2003. An example is the counting iterator: const std::vector<int> values1(boost::make_counting_iterator(1), boost::make_counting_iterator(10)); gives C4244. This is even harder to supress since the STL is included in the precompiled header. Is it an option to have a good look at this issue for the next release? Note that it is also possible that Plauger's STL is not conforming. Wkr, me

5 5